reduce the number of stan models

[sbfnk]

fixes #393 by combining the models from dist_fit into one stan file and ditching the (unused) tune_inv_gamma model (which could of course be brought back if needed).

[seabbs]

tune_inv_gamma model (which could of course be brought back if needed).

I assume you have checked this isn't used?

[sbfnk]

I assume you have checked this isn't used?

Yes - I'm not sure it was ever used, at least I couldn't find a use traversing through git blames.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 5abb62f is merged into main:

•   :ballot_box_with_check:default: 1.06m -> 59.3s [-25.51%, +12.31%]
•   :ballot_box_with_check:no_delays: 53.8s -> 48.9s [-18.3%, +0.08%]
•   :ballot_box_with_check:random_walk: 16.3s -> 16.8s [-7.89%, +14.68%]
•   :ballot_box_with_check:stationary: 40.1s -> 44.2s [-0.35%, +20.55%]
•   :ballot_box_with_check:uncertain: 1.13m -> 1.09m [-11.66%, +4.24%]
  Further explanation regarding interpretation and methodology can be found in the documentation.

[seabbs]

This looks good. I think we need a clean depreciation notice here so ideally that would use lifecycle to report that tune_inv_gamma has been deprecated if called. Even more ideally this could give a link to a gist etc where the code still exists (or the git history) and a note saying to let the authors know if using.

Otherwise this is good to go once it has a news update and a Dev version update.

Nice to speed up package compilation!

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 7215bad is merged into main:

•   :ballot_box_with_check:default: 43.7s -> 49.1s [-3.38%, +28.06%]
•   :ballot_box_with_check:no_delays: 39.9s -> 38.7s [-10.51%, +4.31%]
•   :ballot_box_with_check:random_walk: 12.2s -> 12.5s [-6.2%, +11.12%]
•   :ballot_box_with_check:stationary: 29.8s -> 31s [-6.12%, +14.17%]
•   :ballot_box_with_check:uncertain: 48.1s -> 47s [-17.63%, +13.24%]
  Further explanation regarding interpretation and methodology can be found in the documentation.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 2023b9f is merged into main:

•   :ballot_box_with_check:default: 1.03m -> 1.04m [-12.31%, +14.6%]
•   :ballot_box_with_check:no_delays: 47.9s -> 50.4s [-8.68%, +19.11%]
•   :ballot_box_with_check:random_walk: 16.1s -> 19.2s [-33.71%, +73.41%]
•   :ballot_box_with_check:stationary: 40.9s -> 39.7s [-17.85%, +12.16%]
•   :ballot_box_with_check:uncertain: 59.5s -> 1m [-21.34%, +23.16%]
  Further explanation regarding interpretation and methodology can be found in the documentation.

[seabbs]

LGTM.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 67de94b is merged into main:

•   :ballot_box_with_check:default: 47.9s -> 47.8s [-15.94%, +15.33%]
•   :ballot_box_with_check:no_delays: 40.1s -> 38.8s [-10.63%, +4.03%]
•   :ballot_box_with_check:random_walk: 13s -> 12.1s [-13.84%, +0.01%]
•   :ballot_box_with_check:stationary: 30.1s -> 29.8s [-12.64%, +10.87%]
•   :ballot_box_with_check:uncertain: 48.8s -> 52.5s [-5.56%, +20.68%]
  Further explanation regarding interpretation and methodology can be found in the documentation.